One of the nice things about the WD MyBook is that the embedded Linux is easy enough to update, so installing an SSH daemon was one of the first things I did, and I’m so glad I did. The first sign of trouble was when the network shares on the MyBook suddenly became “read only”, or just disappeared. At first I thought perhaps the NAS device was overheating, but it was no hotter than usual and none of the vents were blocked. I could log in via SSH and shell around the system OK. The details under /var/log showed an interesting story. There was plenty of evidence that the hard drive was failing. Rebooting sometimes restored functionality for an hour or so. I even popped the device into freezer for 20 minutes, which seemed to make it stable for a little while longer. I knew, however, that I probably had less than 3 days before the device would completely collapse.
With SAMBA breaking down due to the recurring disk faults, I decided to rely on secure file transfer over SSH. The encryption overhead put more strain on the CPU, generating a bit more heat, so I dismantled the device and provided better air cooling. Over two days, through countless reboots, I managed to move 350Gb to my main machine. Finally, the device completely refused to respond, so I declared it dead.
So what about my backups? Yes, I had backups of almost everything that was on the failing hard drive, but unfortunately I don’t have a good catalogue. So finding out where the latest version of everything was located would be a painful task. The dying NAS had the latest copy of everything in one place, so making a complete copy of the NAS before it died would be far easier.
If I get my hands on another (non-WD) SATA drive, I might rebuild the device, just for fun. Meanwhile there’s a replacement terabyte NAS on the way. Also non-WD. I’ve learned that lesson.
Categorised as: Uncategorized